Preprocess claim, claimant and article contents

[mattesko]

Acceptance Criteria:

• Apply preprocessing techniques seen in class: Stemming, lemmatizing, removing stopwords
• Apply additional preprocessing techniques: Removing punctuation, replace numbers with word equivalents, lowercase
• Script can be used as a script or as a module
• If used as a script: Preprocess metadata_articles dataframe, concatenate feature columns into one feature column, and pickle and store resulting dataframe under data
• If used as a module: Offers public functions (preprocess, create dataframe for training)